Sensitive content

This media may contain sensitive content.

Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

~ I somehow ended up as a support... Ɛ> High quality version on P-smth <3 Shion Model: dzooworks🔞 Male model: IceDev VA: Misuzugon 🦕 SFX: OpenNSFW 🟣 Available Now #overwatch

DieScope

28,332 subscribers

122,950 görüntüleme • 25 gün önce •via X (Twitter)

NSFW #overwatch

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

Officer Qingyi is chasing down a group of vigilantes who stole Random Play's HDD! She managed to catch up to them but they had the leg up on her... SFX OpenNSFW 🟣 Available Now Qingyi & Zhu Yuan VA Misuzugon 🦕 Even Spicier Version on >> <<

Sensitive content

Officer Qingyi is chasing down a group of vigilantes who stole Random Play's HDD! She managed to catch up to them but they had the leg up on her... SFX OpenNSFW 🟣 Available Now Qingyi & Zhu Yuan VA Misuzugon 🦕 Even Spicier Version on >> <<

ves 🔞

44,065 görüntüleme • 1 yıl önce

Low quality engagement post This is how each Girl in Overwatch farts on a day-to-day basis. Not after a heavy meal or a night of drinking. These are baseline~ In any case, I wanna hit these ladies fumes like a geek bar! SFX by Lil' 🍊 and OpenNSFW 🟣 Available Now Now, if you'll excuse me, I'm going to go goon to some more fart comps. #Overwatch #meme #fart #fartfetish #girlsfarting #eproctophilia #femalefarts

Low quality engagement post This is how each Girl in Overwatch farts on a day-to-day basis. Not after a heavy meal or a night of drinking. These are baseline~ In any case, I wanna hit these ladies fumes like a geek bar! SFX by Lil' 🍊 and OpenNSFW 🟣 Available Now Now, if you'll excuse me, I'm going to go goon to some more fart comps. #Overwatch #meme #fart #fartfetish #girlsfarting #eproctophilia #femalefarts

🩵🩵Seat Shark🩵🩵

28,473 görüntüleme • 2 ay önce

She's here! Tachy's Model is finally done and ready 🎉 I hope everyone likes the avi and model! As always, you can find the Free SFW public versions on my avatar world, or the full version on my PayHip acc Thank you everyone for all the amazing support <3 #Tachy #VRChat

Jaller Arts (Comms Closed)

164,468 görüntüleme • 5 ay önce

GLM 4.6 fans! just soared to the top as the fastest provider in Artificial Analysis for the model. > a 114 TPS and <0.18s TTFT. > That's 2x faster than the next best option on both metrics. Available now in Cline.

GLM 4.6 fans! just soared to the top as the fastest provider in Artificial Analysis for the model. > a 114 TPS and <0.18s TTFT. > That's 2x faster than the next best option on both metrics. Available now in Cline.

Cline

31,584 görüntüleme • 10 ay önce

P-Image-Ideogram, built with Ideogram, now has a Very High mode. Very High uses test-time scaling to push final image quality further. Built for hero assets, difficult prompts, and quality-critical generations when people and products have to pop. • ~5.55s at 1K · $0.033/image (1K) · $0.066/image (2K) • Five modes from Very Low ($0.003) through Very High Again, benchmarks agree: This new very high mode is on efficiency-quality based on human and automated evals from Datapoint AI, Design Arena, and P-Judger. Full benchmarks results follow in a later post on P-Bench. Available on Cloudflare ComfyUI Gamma inference shell Kittl Leonardo.Ai LovartAI Magnific Picsart Replicate Runware Scenario Together AI WaveSpeedAI Wiro AI — Ship AI Faster 👉 Free Playground: 🧩 API: 📚 Model page:

P-Image-Ideogram, built with Ideogram, now has a Very High mode. Very High uses test-time scaling to push final image quality further. Built for hero assets, difficult prompts, and quality-critical generations when people and products have to pop. • ~5.55s at 1K · $0.033/image (1K) · $0.066/image (2K) • Five modes from Very Low ($0.003) through Very High Again, benchmarks agree: This new very high mode is on efficiency-quality based on human and automated evals from Datapoint AI, Design Arena, and P-Judger. Full benchmarks results follow in a later post on P-Bench. Available on Cloudflare ComfyUI Gamma inference shell Kittl Leonardo.Ai LovartAI Magnific Picsart Replicate Runware Scenario Together AI WaveSpeedAI Wiro AI — Ship AI Faster 👉 Free Playground: 🧩 API: 📚 Model page:

Pruna AI

10,077 görüntüleme • 13 gün önce

Here's a look behind the scenes of how I used a 3d model, generated from an image, as a "driver" for ai animation. In the near future we'll see more 3d tools that support these types of workflows, with much more emphasis on enhancing user input instead of just letting the ai do everything. I used the 3d model to generate keyframes that I then animated in Luma Dream. By using a 3d model with a diffusion "layer" done with Krea, I got quite high quality frames with a relatively high consistency as well since the AI didn't have to "hallucinate" that much. I upscaled and detailed the frames using Magnific. Using semi-detailed 3d models as drivers for gen ai is very powerful, and it allows us to use models that are sculpted in a more free-flowing type of workflow that doesn't rely so much on high surface detailing or very time-consuming finish, but instead relies more on gestural 3d sculpting. #art #ai

Here's a look behind the scenes of how I used a 3d model, generated from an image, as a "driver" for ai animation. In the near future we'll see more 3d tools that support these types of workflows, with much more emphasis on enhancing user input instead of just letting the ai do everything. I used the 3d model to generate keyframes that I then animated in Luma Dream. By using a 3d model with a diffusion "layer" done with Krea, I got quite high quality frames with a relatively high consistency as well since the AI didn't have to "hallucinate" that much. I upscaled and detailed the frames using Magnific. Using semi-detailed 3d models as drivers for gen ai is very powerful, and it allows us to use models that are sculpted in a more free-flowing type of workflow that doesn't rely so much on high surface detailing or very time-consuming finish, but instead relies more on gestural 3d sculpting. #art #ai

Martin Nebelong

48,280 görüntüleme • 2 yıl önce

Introducing Soniox TTS v2, our most powerful text-to-speech model yet. Soniox TTS v2 brings extraordinary voice quality, expressive control through audio tags, exceptional precision, high-fidelity voice cloning, more than 60 languages, natural language mixing, and low-latency streaming together in one model. Built from the ground up with a new model architecture, audio codec, and highly optimized inference engine, Soniox TTS v2 delivers frontier TTS at just $0.70 per generated hour. Available globally today as tts-rt-v2. Try it out: Turn the sound on and hear what’s possible. 🔊

Introducing Soniox TTS v2, our most powerful text-to-speech model yet. Soniox TTS v2 brings extraordinary voice quality, expressive control through audio tags, exceptional precision, high-fidelity voice cloning, more than 60 languages, natural language mixing, and low-latency streaming together in one model. Built from the ground up with a new model architecture, audio codec, and highly optimized inference engine, Soniox TTS v2 delivers frontier TTS at just $0.70 per generated hour. Available globally today as tts-rt-v2. Try it out: Turn the sound on and hear what’s possible. 🔊

Soniox

663,421 görüntüleme • 29 gün önce

“The quality is consistently high, not occasionally high. Against Wan2.7, it's not just a huge update, it's like a totally new model. Anyone who’s tried any version of Wan before should retry this one." Lovis Odin put Wan3.0 through its paces across T2V, I2V, and R2V — and went beyond testing the model to share his own prompt techniques for getting the most out of it. From exploring Wan3.0’s capabilities to learning how to write better prompts in the era of 30-second video generation, this session is packed with practical insights for creators. Wan3.0 is now available on fal. Try it on fal ai or fal Agent.

“The quality is consistently high, not occasionally high. Against Wan2.7, it's not just a huge update, it's like a totally new model. Anyone who’s tried any version of Wan before should retry this one." Lovis Odin put Wan3.0 through its paces across T2V, I2V, and R2V — and went beyond testing the model to share his own prompt techniques for getting the most out of it. From exploring Wan3.0’s capabilities to learning how to write better prompts in the era of 30-second video generation, this session is packed with practical insights for creators. Wan3.0 is now available on fal. Try it on fal ai or fal Agent.

Wan

1,839,419 görüntüleme • 16 gün önce

𝗥𝘂𝗻 𝗚𝗼𝗼𝗴𝗹𝗲'𝘀 𝗚𝗲𝗺𝗺𝗮 𝟰 + 𝗢𝗽𝗲𝗻𝗖𝗹𝗮𝘄 𝗮𝘀 𝗮 𝗳𝗿𝗲𝗲 𝗽𝗿𝗶𝘃𝗮𝘁𝗲 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁 𝗼𝗻 𝘆𝗼𝘂𝗿 𝗼𝘄𝗻 𝗺𝗮𝗰𝗵𝗶𝗻𝗲 𝗶𝗻 𝟯 𝘀𝘁𝗲𝗽𝘀. No API bills. No usage limits. No subscription. Nothing leaves your computer. Here's the full setup: → Step 1: Go to Download and install. Update to version 0.2.2.0 or higher. → Step 2: Open terminal. Type: ollama pull gemma4. Downloads the model. Done. → Step 3: Install OpenClaw. Select Ollama as your provider. Point it at port 11434. Pick Gemma 4. That's it. Your AI agent is now running locally. Message it through Telegram, Slack, Discord, or WhatsApp like a coworker. Read files. Write code. Remember context across every conversation. All on your own hardware. Gemma 4 ranked number 3 on the global open model leaderboard on launch day. It beat models with 20 times more parameters. The 26B version activates only 4B parameters at a time so you get near large-model quality at small-model speed. Every AI subscription you're paying for right now could be replaced with this.

𝗥𝘂𝗻 𝗚𝗼𝗼𝗴𝗹𝗲'𝘀 𝗚𝗲𝗺𝗺𝗮 𝟰 + 𝗢𝗽𝗲𝗻𝗖𝗹𝗮𝘄 𝗮𝘀 𝗮 𝗳𝗿𝗲𝗲 𝗽𝗿𝗶𝘃𝗮𝘁𝗲 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁 𝗼𝗻 𝘆𝗼𝘂𝗿 𝗼𝘄𝗻 𝗺𝗮𝗰𝗵𝗶𝗻𝗲 𝗶𝗻 𝟯 𝘀𝘁𝗲𝗽𝘀. No API bills. No usage limits. No subscription. Nothing leaves your computer. Here's the full setup: → Step 1: Go to Download and install. Update to version 0.2.2.0 or higher. → Step 2: Open terminal. Type: ollama pull gemma4. Downloads the model. Done. → Step 3: Install OpenClaw. Select Ollama as your provider. Point it at port 11434. Pick Gemma 4. That's it. Your AI agent is now running locally. Message it through Telegram, Slack, Discord, or WhatsApp like a coworker. Read files. Write code. Remember context across every conversation. All on your own hardware. Gemma 4 ranked number 3 on the global open model leaderboard on launch day. It beat models with 20 times more parameters. The 26B version activates only 4B parameters at a time so you get near large-model quality at small-model speed. Every AI subscription you're paying for right now could be replaced with this.

Julian Goldie SEO

36,898 görüntüleme • 5 ay önce

Guinaifen Possessed During Stream~ 👻 After wrapping up Stream Guinaifen is possessed by some suspicious malware on her phone... SFX OpenNSFW 🟣 Available Now Moan Pack MOO🩷COW This animation is shorter than my other ones due to me testing some new plugins I got for Koikatsu and I wanted to time myself on how long it takes start to finish to make an animation, so I could more accurately price them. More info soon on my Twitter. 4K60, Nude, POV, Phone Camera, No Overlay Version, and Scene Data on Patreon! Discord #TSF #憑依 #乗っ取り

Sensitive content

Guinaifen Possessed During Stream~ 👻 After wrapping up Stream Guinaifen is possessed by some suspicious malware on her phone... SFX OpenNSFW 🟣 Available Now Moan Pack MOO🩷COW This animation is shorter than my other ones due to me testing some new plugins I got for Koikatsu and I wanted to time myself on how long it takes start to finish to make an animation, so I could more accurately price them. More info soon on my Twitter. 4K60, Nude, POV, Phone Camera, No Overlay Version, and Scene Data on Patreon! Discord #TSF #憑依 #乗っ取り

ves 🔞

45,176 görüntüleme • 1 yıl önce

🧵1/3 I partnered with Kling to make a promo for their new 3.0 model. I came up with the concept, created it and delivered it all on my own in 3 days of early access, I wanted to make something that showed how Kling could be used to tell a diverse range of stories in a diverse range of styles. I've honestly been blown away be how incredible the model is and I haven't even scratched the surface of features like the new 3.0 OMNI model. For me Kling 2.6 was already best in class for most things but 3.0 sees massive improvements to video quality and detail (no need to add grain to mask imperfections) and prompt adherence. Also you can time your generations to be as long as 15 seconds and as short as 3, this is fantastic for pacing your scenes. Read on to find out more about the new Performance and Multi shot features!👇

🧵1/3 I partnered with Kling to make a promo for their new 3.0 model. I came up with the concept, created it and delivered it all on my own in 3 days of early access, I wanted to make something that showed how Kling could be used to tell a diverse range of stories in a diverse range of styles. I've honestly been blown away be how incredible the model is and I haven't even scratched the surface of features like the new 3.0 OMNI model. For me Kling 2.6 was already best in class for most things but 3.0 sees massive improvements to video quality and detail (no need to add grain to mask imperfections) and prompt adherence. Also you can time your generations to be as long as 15 seconds and as short as 3, this is fantastic for pacing your scenes. Read on to find out more about the new Performance and Multi shot features!👇

Uncanny Harry AI

29,194 görüntüleme • 7 ay önce

Anybody need a ride? MiLE HiGH TAXi is available on Steam now! I set out over 3 years ago to make a game players can pick up & put down at any time. Something appealing to scifi fans and retro-gamers alike. Done! I truly hope you enjoy playing much as I enjoyed making it! Cass

Anybody need a ride? MiLE HiGH TAXi is available on Steam now! I set out over 3 years ago to make a game players can pick up & put down at any time. Something appealing to scifi fans and retro-gamers alike. Done! I truly hope you enjoy playing much as I enjoyed making it! Cass

Cass | MiLE HiGH TAXi on Steam, Playstation, Epic

193,870 görüntüleme • 3 yıl önce

Head Petting Asset WIP I forget, this one was requested a while back! I am trying my very best to keep up with weekly releases but honestly I'm struggling to keep up with my own targets. I'm trying to focus more on better quality releases which means the process of my work takes longer, alternating between making assets and working on the customizable vtuber model, and wedding prep as I'm getting married next month. A little overwhelmed but I really appreciate everyones patience and continued support. ♡ #vtuberassets #Live2DWIP #vtuberartist

Head Petting Asset WIP I forget, this one was requested a while back! I am trying my very best to keep up with weekly releases but honestly I'm struggling to keep up with my own targets. I'm trying to focus more on better quality releases which means the process of my work takes longer, alternating between making assets and working on the customizable vtuber model, and wedding prep as I'm getting married next month. A little overwhelmed but I really appreciate everyones patience and continued support. ♡ #vtuberassets #Live2DWIP #vtuberartist

Little Mewn ♡ Customizable Model OUT NOW!

30,567 görüntüleme • 1 yıl önce

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:

Tencent Hy

89,257 görüntüleme • 1 yıl önce

After playing Kingdom Come Deliverance 2 last night, I booted up Starfield with the intention of quickly downloading the recent creation club content. 3 hours later and I was sucked right back into this brilliant game and ended up making a bit more progress on the main game (second playthrough). I've said it before and I'll say it again, love this game. Bethesda RPG and space in one package.....🤌 As an aside, with the latest transformer model, this games already very good image quality is sublime PS, the physics in this game, in low gravity, are brilliant (see vid) #Starfield #pcgaming

After playing Kingdom Come Deliverance 2 last night, I booted up Starfield with the intention of quickly downloading the recent creation club content. 3 hours later and I was sucked right back into this brilliant game and ended up making a bit more progress on the main game (second playthrough). I've said it before and I'll say it again, love this game. Bethesda RPG and space in one package.....🤌 As an aside, with the latest transformer model, this games already very good image quality is sublime PS, the physics in this game, in low gravity, are brilliant (see vid) #Starfield #pcgaming

Peter

92,485 görüntüleme • 7 ay önce

✨ I made a new feature called [ 💃 Photo Shoot Simulator ] which lets you do real life photo shoots with real life models on my site Photo AI Upload clothes to try on, or prompt your scene, and your model will now move and pose for you Then when you hit the [ 📸 ] button, you take photos just like you would in a real life photo shoot All your photos are then available upscaled as 4K high resolution pics with a full commercial license to use in your store, social media, marketing, on ads, etc. No need for a model, photographer, or studio anymore!

✨ I made a new feature called [ 💃 Photo Shoot Simulator ] which lets you do real life photo shoots with real life models on my site Photo AI Upload clothes to try on, or prompt your scene, and your model will now move and pose for you Then when you hit the [ 📸 ] button, you take photos just like you would in a real life photo shoot All your photos are then available upscaled as 4K high resolution pics with a full commercial license to use in your store, social media, marketing, on ads, etc. No need for a model, photographer, or studio anymore!

@levelsio

581,055 görüntüleme • 1 yıl önce

Nvidia has just announced Alpamayo 2 Super, an open 34 billion parameter reasoning vision-language-action model designed to accelerate the development of autonomous vehicles. This new model combines the NVIDIA Cosmos 3 Super reasoning model with a 2 billion parameter diffusion-based action expert model, and is post trained with reinforcement learning. The model can return multiple outputs: future trajectory plans, reasoning traces, grounded answers to questions about the scenes, and auto label generation. The model weights are now available for anyone to download on Hugging Face, and the inference code has been posted to GitHub. Distilled models can be deployed commercially without any further permission from Nvidia, and model outputs carry no license conditions. Automakers can distill down a compact version of this model that can run on the Nvidia computer in the car. Major kudos to Nvidia and Jensen Huang for advancing the state of the industry by releasing this as an open model with permissive licensing. Jensen isn't just paying lip service to the idea of open models, Nvidia is actually contributing to the ecosystem — and it's great for their business, because it helps sell more Thor computers that go in the car. Anyone can go download the model and play with it. If you do, let me know what you think. Personally I think it's so cool that we have open weights models that are this advanced, for anyone to download.

Nvidia has just announced Alpamayo 2 Super, an open 34 billion parameter reasoning vision-language-action model designed to accelerate the development of autonomous vehicles. This new model combines the NVIDIA Cosmos 3 Super reasoning model with a 2 billion parameter diffusion-based action expert model, and is post trained with reinforcement learning. The model can return multiple outputs: future trajectory plans, reasoning traces, grounded answers to questions about the scenes, and auto label generation. The model weights are now available for anyone to download on Hugging Face, and the inference code has been posted to GitHub. Distilled models can be deployed commercially without any further permission from Nvidia, and model outputs carry no license conditions. Automakers can distill down a compact version of this model that can run on the Nvidia computer in the car. Major kudos to Nvidia and Jensen Huang for advancing the state of the industry by releasing this as an open model with permissive licensing. Jensen isn't just paying lip service to the idea of open models, Nvidia is actually contributing to the ecosystem — and it's great for their business, because it helps sell more Thor computers that go in the car. Anyone can go download the model and play with it. If you do, let me know what you think. Personally I think it's so cool that we have open weights models that are this advanced, for anyone to download.

Whole Mars Catalog

45,595 görüntüleme • 1 ay önce

🌘 Meet Kimi K2.7 Code HighSpeed! A high-speed mode of our latest open-source multimodal coding model, Kimi K2.7 Code. ⚡️ Up to 6× faster: Around 180 tok/s on coding tasks with median-length inputs, and up to 260 tok/s on shorter-context tasks. 🔷 Rolling out to Kimi Code Beta Program members, Kimi API developers, and Kimi Business users. (Access will remain limited for now due to capacity constraints.) 🔷 No invite needed. Anyone who joins the Beta Program has a chance to get access 👉 Open intelligence should be instant, affordable, and borderless. We'll continue improving the model and expanding access as more capacity becomes available! 🔗 Kimi Code: 🔗 API:

🌘 Meet Kimi K2.7 Code HighSpeed! A high-speed mode of our latest open-source multimodal coding model, Kimi K2.7 Code. ⚡️ Up to 6× faster: Around 180 tok/s on coding tasks with median-length inputs, and up to 260 tok/s on shorter-context tasks. 🔷 Rolling out to Kimi Code Beta Program members, Kimi API developers, and Kimi Business users. (Access will remain limited for now due to capacity constraints.) 🔷 No invite needed. Anyone who joins the Beta Program has a chance to get access 👉 Open intelligence should be instant, affordable, and borderless. We'll continue improving the model and expanding access as more capacity becomes available! 🔗 Kimi Code: 🔗 API:

Kimi.ai

571,391 görüntüleme • 2 ay önce

SGLang now supports DSpark, enabling confidence-driven, variable-length verification for speculative decoding 🎉 DSpark addresses a key bottleneck under load: instead of verifying every draft token, it verifies only where the draft model is confident, so the gains hold even as batch size scales. We heavily optimized variable-length verification in SGLang. Across batch sizes 1 to 256, DSpark gives the best throughput/latency tradeoff on DeepSeek-V4-Flash, ahead of both MTP and non-spec. At high concurrency, dynamic scheduling provides up to ~20% higher throughput compared to a fixed budget, while maintaining high verification quality across workloads. With fused kernels and zero-overhead scheduling, DeepSeek-V4-Pro reaches 383.7 tok/s at B=1 on B300. DSpark is now available in SGLang with support for Qwen3 and DeepSeek-V4. Thanks DeepSeek for open-sourcing! Blog with full technical details and commands to run below 👇

SGLang now supports DSpark, enabling confidence-driven, variable-length verification for speculative decoding 🎉 DSpark addresses a key bottleneck under load: instead of verifying every draft token, it verifies only where the draft model is confident, so the gains hold even as batch size scales. We heavily optimized variable-length verification in SGLang. Across batch sizes 1 to 256, DSpark gives the best throughput/latency tradeoff on DeepSeek-V4-Flash, ahead of both MTP and non-spec. At high concurrency, dynamic scheduling provides up to ~20% higher throughput compared to a fixed budget, while maintaining high verification quality across workloads. With fused kernels and zero-overhead scheduling, DeepSeek-V4-Pro reaches 383.7 tok/s at B=1 on B300. DSpark is now available in SGLang with support for Qwen3 and DeepSeek-V4. Thanks DeepSeek for open-sourcing! Blog with full technical details and commands to run below 👇

LMSYS Org

170,303 görüntüleme • 2 ay önce

SAKANA FUGU ULTRA vs. CLAUDE OPUS 4.8 RESULTS Prompt: "build a really high quality single html file crossy road game with three.js" Sakana Fugu Ultra: - Tokens Used: ~89k ($7.32) - Time Elapsed: 22 minutes - Issues: inverted directional turn, wonky camera, no sfx, not identical to Crossy Road game Claude Opus 4.8 Ultracode: - Tokens Used: ~940k (~$37.85) - Time elapsed: 79 minutes - Issues: got stuck twice in a retry loop (had to prompt to self-correct), wrong character position after restart, difficult from the start (whereas Fugu's version got significantly more difficult as you progress, which is the correct behaviour) I think in terms of application functionality, quality, and design, Opus won. In terms of model speed and performance, Fugu on Opencode won. What are your thoughts? Comment down below who you think won! ⬇️

SAKANA FUGU ULTRA vs. CLAUDE OPUS 4.8 RESULTS Prompt: "build a really high quality single html file crossy road game with three.js" Sakana Fugu Ultra: - Tokens Used: ~89k ($7.32) - Time Elapsed: 22 minutes - Issues: inverted directional turn, wonky camera, no sfx, not identical to Crossy Road game Claude Opus 4.8 Ultracode: - Tokens Used: ~940k (~$37.85) - Time elapsed: 79 minutes - Issues: got stuck twice in a retry loop (had to prompt to self-correct), wrong character position after restart, difficult from the start (whereas Fugu's version got significantly more difficult as you progress, which is the correct behaviour) I think in terms of application functionality, quality, and design, Opus won. In terms of model speed and performance, Fugu on Opencode won. What are your thoughts? Comment down below who you think won! ⬇️

Mark Santos

685,792 görüntüleme • 2 ay önce